It has been shown by us that valine is incorporated into isoleucine specified sites in rabbit hemoglobin about one in 3000 times. Improving and extending this experimental approach we propose to determine the frequency with which several other amino acids are inadvertently substituted for the genetically correct amino acid and to relate such mistranslations to similarities in genetic code or similarities in chemical structure of the amino acid involved. In parallel with the above, we propose to develop tecnniques for achieving high levels of incorporation of natural (3H) amino acids into albumin and ferritin, to isolate the proteins, to cleave them into a limited number of well-defined peptides. Using techniques similar to those developed for hemoglobin we plan to measure a variety of translation errors and then to determine whether these errors become more frequent with the age of the rat as would be expected from the Orgel hypothesis of aging.